ADP: Private Sector Added Just 38,000 Jobs in August

Thirty-eight thousand. That's how many jobs private employers added in August, according to the ADP National Employment Report published Wednesday morning, and education and health services accounted for 45,000 of them.

Strip out schools and hospitals and August was a net job loss

ADP Research, which produces the report with the Stanford Digital Economy Lab, said private-sector employment rose by 38,000 in August. Economists polled by Bloomberg had expected 47,000, according to Yahoo Finance. Fox Business put the consensus at 48,000. It's the same miss either way: different economist panels, roughly a 10,000-job shortfall, and a second straight month of hiring that barely registers.

The mix matters more than the headline. Service-providing industries added 48,000 jobs while goods producers lost 10,000, with manufacturing shedding 17,000 and construction adding 12,000, according to Crypto Briefing's breakdown of the ADP data. Education and health services delivered 45,000 on their own. Do the subtraction and everything outside classrooms and clinics went backwards.

That is a very narrow economy.

July was revised up to 46,000 from an initially reported 44,000, and August marks the slowest pace of private hiring since January, Crypto Briefing reported. Firms with 500 or more employees accounted for 34,000 of the net gains, so what hiring there is sits with the biggest employers. Small businesses aren't driving this.

The number worth watching is the 1.7-point pay gap between movers and stayers

ADP's own release put base pay growth at 3.2% year over year for all private-sector workers and gross pay growth at 4.7%. Split it by mobility and the picture sharpens: base pay for job-stayers rose 3.0%, for job-changers 4.7%. On gross pay, stayers got 4.4% and changers got 7.3%.

So is it still worth moving? On the arithmetic, yes, and by a wide margin, but the catch is obvious once you look at where the jobs actually are. A 1.7-point base-pay premium for switching is only collectable if someone in your industry is hiring, and in August manufacturing and business services were cutting. For everyone staying put, 3.0% base pay growth against average hourly earnings running at 3.2% in the BLS data means real income is going roughly nowhere.

ADP says slow. The BLS says worse than slow.

These two series have been diverging, and the official one is the grimmer read. Total nonfarm payroll employment fell by 23,000 in July, the BLS reported, against an average monthly gain of 34,000 over the prior 12 months. Financial activities lost 14,000 jobs in July and are down 121,000 from a peak in May 2025. Health care added 22,000, well below its 36,000 monthly average over the previous year, which is the same warning ADP's August print carries: the one reliable engine is slowing too.

ADP isn't a forecast of the BLS number and never has been. It's payroll records for more than 26 million employees, measured on a different reference basis and covering private employers only, so it misses the government swings that dominated July. Anyone treating Wednesday's 38,000 as a firm prediction for Friday is selling something.

What it does tell you is direction, and the direction has been down for three months.

No, there's no ADP cyber attack behind Wednesday's weak print

Every time ADP tops the wires, searches for "adp cyber attack update" and "adp cyber security" climb with it. Nothing in this release points to a security incident, an outage or a data problem. ADP describes the report as built on anonymized, aggregated payroll data from over 26 million US employees, and the August publication went out on its normal 8:15 a.m. ET schedule with the pay analytics expanded to 56 metropolitan areas. The soft number is a labor market story, not an IT one. A payroll processor of this scale is a permanent target, and if that ever changed the tell would be in the release notes, not in the seasonal adjustment.

Four dates decide whether 38,000 was noise

• Friday, September 4: the BLS August Employment Situation, the number that actually moves the bond market. After July's 23,000 decline, a second negative print changes the conversation entirely.
• Friday, September 11: August CPI. Inflation is still above the Fed's 2% goal, and Brent crude near $89.70 a barrel isn't helping the energy component, per GO Markets' September outlook.
• September 15-16: the FOMC meeting, with the decision at 2 p.m. ET on the 16th and an updated Summary of Economic Projections. The target range has been parked at 3.50% to 3.75%.
• Wednesday, September 30: PCE inflation, the Fed's preferred gauge, and the last major data point of the quarter.

Who wins from a labor market this weak? Borrowers, eventually. Softer payrolls pull the front end of the curve lower and drag mortgage and refinancing costs with it, and anyone with a rate reset due in 2027 should quietly want more prints like this one. Savers lose the mirror image: deposit and money-market yields track the same expectations down, and the 5% headline rates of the recent past aren't coming back if hiring keeps stalling.

The losers are more specific than that. Anyone in manufacturing or professional services looking for work is competing in sectors that shrank in August, and the pay premium for switching jobs, real as it is, doesn't help if nobody's advertising. Consumer-facing companies should be reading the pay split closely: gross pay growth of 4.4% for workers who stay put is thin cover for another year of price rises.

One number to keep in your head through Friday. Take out education and health, and America's private sector didn't add a single net job last month.